1f52b on Nostr: Been thinking about this a bit -- *perhaps*, one of the reasons Lightning has taken ...

Been thinking about this a bit -- *perhaps*, one of the reasons Lightning has taken so long is that we're only just realising how important the LSP component is. Wallets/nodes that work with a LSP are so much easier to run and use for non-dev and non-in-it-as-a-business people. Even as a dev, I'm fairly hesitant to sign myself up for the maintenance effort of running a normal Lightning node.

Things like Phoenix, ZEUS (npub1xnf…lpr5), Phoenixd and OpenSecret (npub1mut…23vg) (mutiny server soon! 👀) are just so much easier to work with than "just" run LND/Eclair/C-Lightning/LDK-Node etc. Is the logical conclusion that for Lightning node-running to become prolific, LSPs (or more accurately, a way for nodes to 'buy' inbound from and get channels with others) have to become an inherent part of the Lightning protocol? Feels like it to me.

quoting
nevent1q…67jn

Let’s be clear, if no one steps up and decides to run LSPs because everyone is worried about regulatory concerns, every bitcoin L2 system is toast - every one that has a reasonable security model relies on some kind of centralized or federated party that has similar concerns, even if they can’t seize funds.

Without any L2 systems everyone using bitcoin will simply use custodial platforms because that’s the only way to get reasonable fees and payment latency.

And don’t go yelling at ACINQ for deciding not to operate Phoenix in the US, the software required to run an LSP is open source, with only relatively minimal liquidity allocation logic required to get started. We need new entrants, and that means new companies who think the risk is manageable (I’m confident it is, but I can’t fault anyone for not wanting to take that risk).

If you see someone suggesting ACINQ should just keep running, the correct response is “well why aren’t you running an LSP”.